OK, I quickly released an update. I had no real time to test all platforms, so please only update if you really experienced a bug. If others confirm it's OK, it might be safe.
Sorry - but I'm really busy these days.
Not really a bug but...
Downloading an update is initiated with the "Start Update", once pressed the update starts downloading but the button "Start Update" remains on the dialogue, so a noobie will almost always be tempted to click on it again to make sure it was clicked. This causes errors and a final crash. Possibly disabling the button or replacing it with a "Cancel" button would be better?
Cheers,
Ian
P.S. Also, this happens, ever so often, if my firewall asks to allow the EditorE.exe to access the net and I allow it, the pause to do so, seems to cause EditorE.exe to hang in memory. I have to cancel the process and re-run GLB. This is an old old old... bug.
Yes, I've mentioned that a while ago :)
The editor tries to find an update then. Does that not work?
SAVESPRITE is not saving properly. The dimensions are there but no image png/bmp. :( Please confirm!
Due to my current project, I've uninstalled 8.1 and reverted to 8.0. (and SAVESPRITE works fine).
There is a sprite, but its rather wrong :)
Thanks for the corrected confirmation MrTAToad. :coke:
For some reason my one comes out as purple if you save as PNG and black if its BMP... :blink:
Sorry for the slow reply Gernot,
The first, "Start Update", is only a problem if you press it twice during the update.
The second, only affects new users with firewalls, for a first installation or a full download update.
Just small things, people could get confused with.
Cheers,
Ian
The button should really be disabled once pressed, with it being re-enabled if there was a problem or cancelled in some way :)
Not sure how to explain this bug, but it only started the second after I had updated. Which I`d read the forum b4 I did an update tho, I only updated hoping to fix the bug with comments in folded functions. Anyways, I have an update func so I can use it for any screen size, which worked fine (I even had it running in 64x32 pixels at one point ;) ). But after the update this appears to be screwy, and the bottom part appears to be corrupt (see image)
the code I use for this is ....
FUNCTION Update:
Show_FX()
Show_Particles()
LOCAL sx,sy
GETSCREENSIZE sx,sy
dtime = GETTIMER()
fps = ((1000/dtime)+fps)/2
LOCAL fp=INTEGER(fps)
USESCREEN -1
STARTPOLY 0
POLYVECTOR 0,0,0,0
POLYVECTOR sx,0,800,0
POLYVECTOR sx,sy,800,600
POLYVECTOR 0,sy,0,600
ENDPOLY
PRINT "FPS:"+fp,0,sy-32
SHOWSCREEN
USESCREEN 0
DRAWSPRITE 1000,0,0
ENDFUNCTION
Im pretty sure this isnt me as, like I say, its been working for different resolutions for the past week, but after the update no longer works. It works so long as the screen res is greate than my working screen size. Same thing also happens in full screen as well.
Also, is it possible to rollback an update ? would be pretty handy if it could be done.
[attachment deleted by admin]
It does look like Update is a bit inefficient, especially if you are continually getting the screensize. Why it should be blurry I dont know, but it does look like a previous screen is bleeding into the currently displayed one.
However, its certainly possible to revert back to a previous version - assuming you set the update to backup the original files...
If so, all you need to do is unzip them back into the program directory (making sure that you close GLBasic first). If you dont automatically back up files, then the best way would be find someone who has :)
when is the grabsprite bug going to be fix'd?
YIKES!
Minion - can you make a very small sample that I can use to test the bug?
I'll investiate the SAVESPRITE
What's the problem with GRABSPRITE, please?
i mean savesprite is purpurle i was thinking it was grabsprite and savesprite where the same :)
sorry
Gernot, some code for ya.
Any screen sizes below that of the CREATESCREEN size screws up (either X or Y)
Anything equal or above is fine.
SETSCREEN 640,480,0
//SETSCREEN 800,600,0
//SETSCREEN 1280,1024,0
//SETSCREEN 850,850,0
//SETSCREEN 800,300,0
//SETSCREEN 600,600,0
LOCAL sx,sy
GETSCREENSIZE sx,sy
CREATESCREEN 0,0,800,600
USESCREEN 0
FOR m=0 TO 800 STEP 10
DRAWLINE m,0,0,600,RGB(255,255,255)
DRAWLINE 800-m,600,800,0,RGB(255,255,255)
NEXT
REPEAT
USESCREEN -1
STARTPOLY 0
POLYVECTOR 0,0,0,0
POLYVECTOR sx,0,800,0
POLYVECTOR sx,sy,800,600
POLYVECTOR 0,sy,0,600
ENDPOLY
SHOWSCREEN
USESCREEN 0
UNTIL 0
I have had to go back, as new editor mess my SUBs list on the right window. Will do a screenshoot later.
There appears to be a new update (8.12500), have these issues been addresses or is this for other problems ?
Eeeep ! Another problem ! Just trird to update my version of GLB on my lappy b4 I go on hols (Im taking lappy with me). The Update states that there is a new update .... Vers 8.00000 and my old version is ver 7.34100. When I try to update this the updater crashes ;( Any reason why ? Is there a way to get the update manually ?
SAVESPRITE works fine with the update! =D
QuoteEeeep ! Another problem ! Just trird to update my version of GLB on my lappy b4 I go on hols (Im taking lappy with me). The Update states that there is a new update .... Vers 8.00000 and my old version is ver 7.34100. When I try to update this the updater crashes ;( Any reason why ? Is there a way to get the update manually ?
You can't update from V7.XX to V8.XX you have to install the new version 8, then update - download HERE (http://www.glbasic.com/files/glbasic_sdk.exe) approx. 130Mb
Quote from: Minion on 2010-Oct-09
Eeeep ! Another problem ! Just trird to update my version of GLB on my lappy b4 I go on hols (Im taking lappy with me). The Update states that there is a new update .... Vers 8.00000 and my old version is ver 7.34100. When I try to update this the updater crashes ;( Any reason why ? Is there a way to get the update manually ?
Expanding on what Ian said, the updater is only for updates. You can't use it to do upgrades.
First time I noticed; this messes up the jump list when all on one line (have changed all my TYPES to separate lines :puke:):
TYPE objAmmo; x; y; ENDTYPE
[attachment deleted by admin]
Quote from: matchy on 2010-Oct-10
First time I noticed; this messes up the jump list when all on one line (have changed all my TYPES to separate lines :puke:):
TYPE objAmmo; x; y; ENDTYPE
Just separate the last type def. It worked for me.
OK, the big update is live now.
I hope it's not a halloween gift :S
It seems pretty stable. I've been working with it for a few weeks now.
I am in holidays (no computer!!!!). Can someone enumerate the new things?
The weekend is here, so can give it a good thrashing :whistle:
What new is :
// 8.148
// Compiler:
// Later declared TYPEs cause no more warning.
// Member functions returning a type work now.
// LOCAL a[] AS double // works now -> the array was dropped before.
// BREAK/CONTINUE work inside REPEAT...UNTIL.
//
// Core:
// Better error messages with GETLASTERROR$ in:
// LOAD... and OPENFILE
//
// CLEARSCREEN -1 -> does not clear the backbuffer
// on SHOWSCREEN. (Faster if you take care of that).
//
// SHOWSCREEN: Z-Buffer only gets cleared if X_MAKED was used.
//
// iPhone:
// using GL_EXT_BGRA on iOS 3.x and higher -> slightly faster
// LOADSPRITE, SAVESPRITE, GRABSPPRITE and USEASBMP.
//
// Pandora:
// using OpenGL|ES now. Run! Run!
//
// Commands:
// PRINT supports font kerning with an additional parameter.
// SPLITSTR can keep empty tokens now.
//
// Editor:
// Functions within types are coloured keywords now.
// Replace in selection yields some feedback what it did.
//
// iPhone XCode directory will no longer overwrite project
// and plist files.
Can you write it without quotes ? iPhone don't show a scroll bar in quotes/codes :(
use two fingered multi touch to scroll down in the quotes
I think spaces take up a bit too much, well, space...
Works well though
[attachment deleted by admin]
The kerning is cool :good: :good: and setting the margins in the font editor will probably deal with the spacing.
With the CLEARSCREEN -1, everything seems to alternate between the backbuffer and the forground buffer, producing a flickering effect.
DEBUG RGB(255,255,255)
CLEARSCREEN -1
WHILE TRUE
DRAWRECT RND(200),RND(200),RND(200),RND(200),RGB(RND(255),RND(255),RND(255))
SHOWSCREEN
KEYWAIT
WEND
Try it with the KEYWAIT first and you'll see two completely different screens being built. Remove the KEYWAIT and watch it flicker :)
Debug cant be used with the new numeric types :(
This :
LOCAL a[] AS double
DIM a[6]
a[0]=1.2
DEBUG a[0]+"\n"
Produces :
"TestTType.gbas"(6) warning : probably unassigned variable : a+"\n"
"TestTType.gbas"(6) error : variable is not explicitly defined : a+"\n"
Function name checking is a lot tighter, by the looks of things, which is good :) It spotted a function (in an extended type) that had the name of STRETCHSPRITE...
Quote from: MrTAToad on 2010-Oct-29
Debug cant be used with the new numeric types :(
This :
LOCAL a[] AS double
DIM a[6]
a[0]=1.2
DEBUG a[0]+"\n"
Produces :
"TestTType.gbas"(6) warning : probably unassigned variable : a+"\n"
"TestTType.gbas"(6) error : variable is not explicitly defined : a+"\n"
it seems the built in operators do not work for these types properly when using them as an argument.
It seems the "comments inside subs" bug is again in. Thi time no errors, just blank debug window after f5. Anyone can test this?
Quoteit seems the built in operators do not work for these types properly when using them as an argument.
Thats a pity!
QuoteIt seems the "comments inside subs" bug is again in. Thi time no errors, just blank debug window after f5. Anyone can test this?
Have you got any example code ?